1. The LHCb VErtex LOcator Doris Eckstein Universität Hamburg, Institut für Experimentalphysik The LHCb Experiment - Motivation The LHCb Detector The VErtex LOcator - Requirements - Development and Tests - Production and Installation DESY Seminar, 20 November 2007

2. 20 November 2007Doris Eckstein, DESY Seminar2 LHCb ATLAS ALICE CMS LHC Tunnel Geneva CERN > 600 scientists 47 universities and laboratories 15 countries LHCb at the LHC

3. 20 November 2007Doris Eckstein, DESY Seminar3 (0,0) (0,1) (1- 2 /2)( ,  ) V ub V ud * V td V tb * V cd V cb *    (0,0) (,)(,) V ub V tb V cd V cb * * V ud V td V cd V cb * *    (1 -   ,   ) 2 V us V ts V cd V cb * *  At LHCb terms up to 5 must be considered Triangles almost identical, differences are at the per cent level Unitarity Triangle Slide from Malcolm John

4. 20 November 2007Doris Eckstein, DESY Seminar4 Unitarity Triangle - 2007 Need significant constraint on  Need Precision Measurement on  and further decrease errors on  and  Bs mixing phase  s =  2  Currently from direct measurements:

5. 20 November 2007Doris Eckstein, DESY Seminar5 LHCb has a rich physics program and most analyses expect good results in the early period (<2fb  1 ): –  (  ) LHCb  5 degrees from B s  D s ± K ±, B 0  DK etc –  (  s ) LHCb  0.02 radians –Observation of B s →  –Sensitivity to New Physics phase in B s →  In addition, –  (  m s )  0.012ps –1 –  (sin(2  ))  0.02 (2x10 5 /2fb –1 ) [final B-factory result: σ(sin(2  ))  ± 0.017 stat ] –  (  )  10 degrees –Charm physics LCHb Physics goals Expected constraints on Unitarity Triangle after5 years of LHCb data (10 fb-1) if all measurements agree with the Standard Model

6. 20 November 2007Doris Eckstein, DESY Seminar6 LHCb experiment to study CP violation in B-hadron decays At LHC: pp-collisions  s =14TeV Bunch crossing frequency 40 MHz Pile-up at high luminosity  choose 2x10 32 cm -2 s -1  most events have single interactions Beams are locally less focused Eases reconstruction (B-decay vertex) Lower radiation level  can come closer to beam Interactions/crossing The LHC Environment

7. 20 November 2007Doris Eckstein, DESY Seminar7 LHCb – a forward Spectrometer full spectrum of B-hadrons produced LHCb: equippes the forward direction 15-250mrad acceptance ~40% ~10% B-cross-section large ~500μb 10 12 b-hadrons per nominal year [10 7 s] of data taking (2 fb -1 ) b b b b b b Does not occur

8. 20 November 2007Doris Eckstein, DESY Seminar8 p p 250 mrad 10 mrad Tracking system VELO Trigger Tracker Inner/Outer Tracker Particle ID RICH1 and RICH2 Calorimeters Muon system Kinematics Magnet + Trackers Calorimeters Vertex Reconstruction VELO The Detector

9. 20 November 2007Doris Eckstein, DESY Seminar9 LHCb – at Point 8 Muon Calorimeters RICH2 Trackers Magnet RICH1 VELO

10. 20 November 2007Doris Eckstein, DESY Seminar10 Trigger Only ~1% of inelastic collisions produce b-quarks Branching fractions of interesting B decays are <10 -4 L0 High p t hadron, lepton,  Flag multiple interactions, busy events Hardware (custom boards), latency 4  s Calo, Muon, Pileup, SPD HLT Inclusive and exclusive selections Software (PC farm 1800 nodes), complete event Full information from detector On tape 10 MHz visible 1 MHz full detector readout 2 kHz 35 kB per event Output rate Trigger TypePhysics Use 200 HzExclusive B candidates Specific final states 600 HzHigh Mass di- muons J/ , b  J/  X 300 HzD* CandidatesCharm, calibrations 900 HzInclusive b (e.g. b  ) B data mining VELO Information for fast reconstruction  Fast data reduction

11. 20 November 2007Doris Eckstein, DESY Seminar11 VELO requirements Measure proper time of B decay: t = m B L / pc  decay length L (~ 1 cm in LHCb)  momentum p from decay products (which have ~ 1–100 GeV) Tracker: tracking before magnet cover full downstream detector acceptance  21 stations allowing to measure >=3hits/track Number of hits per particle Pseudorapidity

12. 20 November 2007Doris Eckstein, DESY Seminar12 RF foil interaction point Silicon sensors VELO requirements 1m Interaction Region,  5,3 cm Vertex detector: Reconstruct pp interaction vertex < 10μm wide spread of interaction region in z  many stations around z=0 Reconstruct B decay vertex short track extrapolation distance and minimal multiple scattering IP < 40 μm  minimal material before first measured point  VELO sensors as close as possible to beam 7mm distance  no beam pipe

13. 20 November 2007Doris Eckstein, DESY Seminar13 Vacuum Vessel secondary Vacuum box retractable detector halves vacuum feedthroughs Injection: larger aperture required  allow for retraction by 30mm  sensors in detector vacuum 10 -5 mbar Protect sensors against RF pickup from LHC beam Protect the LHC Vacuum from possible outgassing of detector modules  RF boxes In the Vacuum

14. 20 November 2007Doris Eckstein, DESY Seminar14 Middle station Far station an example V dep R/cm V dep extremely inhomogeneous irradiation dependence on R and station (z) 5x10 12 to 1.3x10 14 n eq /cm 2 /year (compatible with other LHC detectors) Maintain a good S/N performance for at least 3 years Extensive R&D program to select sensor and optimize Front-End chip sensors: oxygenated n-on-n need cooling of detector modules Radiation environment

15. 20 November 2007Doris Eckstein, DESY Seminar15 n + -on-n strips Routing lines on 2 nd metal layer n + implants SiO 2 p + implant Depletion fraction resolution p-on-n efficiency degrades fast n-on-n efficiency ~100% for only 60% depletion depth n-on-n silicon, under-depleted: Limited loss in CCE Less resolution degradation

16. 20 November 2007Doris Eckstein, DESY Seminar16 Sensors some more requirements: Fast 2D tracking and vertexing for Trigger motivates R-and Φ-measuring sensors Optimize resolution + occupancy pitch small at inner radii, larger at larger radii Φ-measuring sensor 40-102μm (R Sensor) 36-97μm (Φ Sensor) sensitive area from 8 to 42mm radius 300μm thickness Single sided 2048 channels per sensor 2 x 2 x 21 sensors  172 k channels © PPARC R measuring sensor 2 nd metal layer Pitch adaptor RO chips

17. 20 November 2007Doris Eckstein, DESY Seminar17 Module Double sided (R-sensor + Φ-sensor) Minimal material budget Kapton hybrids on Carbon fibre substrate 0 CTE Cooled by CO 2 cooling Precision of mounting for Trigger and Foil Metrology of Modules Paddle Hybrid Cooling cookies

18. 20 November 2007Doris Eckstein, DESY Seminar18 Module Burnin 45 modules visually inspected – 483,000 bonds 36 modules fully burned in 14 step process including temp cycling, chip burnin, thermal images, vacuum operation

19. 20 November 2007Doris Eckstein, DESY Seminar19 Insert modules Assembly Take an empty half Connect cooling and cables Electrical test to reveal problems Log fingerprint of each module S/N Testpulse-S/N for all R-sensors

20. 20 November 2007Doris Eckstein, DESY Seminar20 The signal chain 2048 x 84 channels produce data need online zero suppression 60m cat6 TELL1 board To PC farm 1MHz of data

21. 20 November 2007Doris Eckstein, DESY Seminar21 Online zero suppression ADC FIR correction Pedestal correction/ Bit limit Reordering LCMS Clustering 10 bit 8 bit 7 bit All on FPGAs Correct for cross talk and cable effects Take into account complex strip reordering (non-consecutive strips on consecutive readout channels) Implement algorithm for correction of common mode Clustering  read out only cluster data Data should serve a fast Trigger as well as sophisticated offline reconstruction purposes  Online cluster position calculated  included in Cluster format as well as all cluster strip info  TELL1 Emulation for bit-perfect offline code development cut tuning, etc.

22. 20 November 2007Doris Eckstein, DESY Seminar22 Online zero suppression TELL1 Algorithm Cluster Seeding Threshold Cluster Finding Efficiency Emulation developed and tested with testbeam and lab data Keep control about what happens online

23. 20 November 2007Doris Eckstein, DESY Seminar23 Cross talk Cross talk varied between 5 and 20% analysis developed and FIR corrections extracted Implemented in alignment with dramatic improvement in residual distributions

24. 20 November 2007Doris Eckstein, DESY Seminar24 Beam z y x VELO System Test – Testbeam Nov’06

25. 20 November 2007Doris Eckstein, DESY Seminar25 Noise and S/N Signal / Noise HP1HP2HP3HP4 CM noise calculated in groups of 32 channels Noise is stable throughout the data taking and is 1.9-2.6 ADC counts for R sensors and 1.7-2.2 ADC counts for f sensors. (1 ADC ~ 500 e-) S/N ~ 23 for R, better for Φ

26. 20 November 2007Doris Eckstein, DESY Seminar26 Reconstructing the Vertex d = 2mm d = 5mm 15 mm VELO retracted by 30mm during injection Moving: reconstruct beam position  move in Iterations Standard (fast) VELO tracking does not work (R-Φ off centre) Special tracking developed Targets installed during testbeam

27. 20 November 2007Doris Eckstein, DESY Seminar27 Reconstructing the Vertex Target Vertex reconstructed from interaction between proton beam (180 GeV) and sensors or targets. before alignment …and after alignment

28. 20 November 2007Doris Eckstein, DESY Seminar28 Sensor Resolution  40 =8.4  m  40 =8.6  m Alignment at sophisticated level –baseline cluster resolution can be extracted –further improvements gained at large pitch by eta corrections

29. 20 November 2007Doris Eckstein, DESY Seminar29 Simulated Event

30. 20 November 2007Doris Eckstein, DESY Seminar30  IP = 14  m+35  m/p T Impact parameter resolution proper time t=l  m/p Proper time resolution is dominated by B vertex resolution Impact parameter resolution crucial for proper time resolution ~40 fs for most channels Bs→DsBs→Ds Detector Performance – Vertexing

31. 20 November 2007Doris Eckstein, DESY Seminar31 From Assembly to Installation

32. 20 November 2007Doris Eckstein, DESY Seminar32 Where it has to fit What the beam sees Should be a perfect match

33. 20 November 2007Doris Eckstein, DESY Seminar33 VELO Installation It was electrically tested and it does work! No damage occured during the installation! Now Commissioning.

34. 20 November 2007Doris Eckstein, DESY Seminar34 Summary LHCb is a dedicated B-Physics experiment at the LHC. The LHCb VELO is a crucial part of the detector and will contribute to reaching the experiments physics goals. The VELO was recently installed and commissioning is ongoing. The commissioning in the testbeam helped to understand the detector and to reach the expected detector performance. Thanks to my former VELO colleagues! Good luck for the data taking enjoy the exciting time ahead!

35. 20 November 2007Doris Eckstein, DESY Seminar35 BACKUP

36. 20 November 2007Doris Eckstein, DESY Seminar36 Tracking T1, T2, T3 made of Outer Tracker and Inner Tracker Trigger Tracker Measurement in fringe field of magnet Covers full detector acceptance  Provide p t for Trigger (together with VELO)  2*2 layers  Silicon microstrip sensors  500  m thickness  ~200  m readout pitch Outer Tracker Inner Tracker for Region of high occupancy cm

37. 20 November 2007Doris Eckstein, DESY Seminar37 Tracking Inner Tracker: only 2% of area, but 20% of tracks  Silicon microstrip sensors 11 cm strips, ~200  m pitch Outer Tracker: 3 stations each made up of 4 double-layers of Kapton/Al straw tubes glued together to form modules two-sensor ladders: 410  m thickness Single sensors: 320  m thickness Tracking behind the Magnet – IT and OT

38. 20 November 2007Doris Eckstein, DESY Seminar38  from B s → D s K  Expect 6200 D s K events in 2 fb –1 B/S < 0.5 Expect 6200 D s K events in 2 fb –1 B/S < 0.5 Expect 140 000 D s  98% suppression achieved with RICH PID system in the analysis Used to measure  m s 2 fb –1 :  (  m s )  0.012ps –1 Expect 140 000 D s  98% suppression achieved with RICH PID system in the analysis Used to measure  m s 2 fb –1 :  (  m s )  0.012ps –1 + ch.c. diagrams Study sensitivity by generating toy-experiments with experimental inputs derived from full MC (Decay time and mass resolution, reconstruction efficiency, tagging…) –Sensitivity with 2 fb -1 : σ(  ) ~ 13° Two tree decays (b  c and b  u), which interfere via B s mixing: –can determine (  s +  ), hence  in a very clean way Fit 4 tagged, time-dependent rates –Extract  s + , strong phase difference , amplitude ratio –B s  D s  also used in the fit to constrain other parameters ( ,  m s,  s ) Slide from Malcolm John

39. 20 November 2007Doris Eckstein, DESY Seminar39 B s mixing phase:  s The equivalent of “sin2  “ for B s mesons In the standard model,  s is small: = -2arg(V ts )  0.036  0.003 –Could be larger if New Physics is present in the box diagram –Recent D0 result  s = –0.79 ±0.56(stat) +0.14–0.01(syst) with 1.1 fb –1 To resolve B s oscillations, excellent proper time resolution is required Modes sensitive to  s : B s → J/  B s →  c  B s → J/   B s → D s D s Control channel (  m s ): B s → D s  Illustration of CPV: toy-modeling LHCb data with  s =  0.2 (i.e. 5  SM) events tagged as B s Slide from Malcolm John

40. 20 November 2007Doris Eckstein, DESY Seminar40 B s  expected sensitivity Very exciting possibility of sensitivity to New Physics enhancement in the early period Current upper limit from the Tevatron is around 20 x SM prediction The dominate background is b , b . –Background analysis is currently limited by Monte Carlo statistics (generation) LHCb’s superior B s invariant mass resolution is crucial in the background rejection LHCb limit on BR at 90% CL (only bkg is observed) LHCb sensitivity (signal+bkg is observed) 5  observation 3  evidence SM BF (x10 –9 ) Integrated luminosity (fb –1 ) BF (x10 –9 ) Expected final CDF+D0 limit SM “early ” perio d Uncertainty in bkg prediction Slide from Malcolm John

41. 20 November 2007Doris Eckstein, DESY Seminar41 Noise inner strip + routingline outer strip Noise (ADC cnts) ΦR inner outer Increasing Strip length Different noise levels understood, primary reason strip geometry and routing Total Noise CM sub. Noise

42. 20 November 2007Doris Eckstein, DESY Seminar42 red = detected hits blue = reconstructed tracks VELO TT T1 T2 T3 Detector Performance – Tracking Momentum resolution  p/p ~ 0.4% tracks passing through full spectrometer:   ~ 95%, a few percent of ghost tracks Mass resolution B s  D s   Typical  m ~15 MeV